The clinical effectiveness of nursing models of diabetes care: A synthesis of the evidence.

To determine the clinical effectiveness (glycemic control, other biological measures, cost-effectiveness and patient satisfaction) of primary care nurse-led interventions for diabetes. A systematic review following methods described for complex interventions and using PRISMA guidelines for reporting was undertaken. Nurse-led care for diabetes can be regarded as a complex intervention requiring the measurement of more than one outcome and for this reason we chose a range of outcomes clinical (symptoms), patient-centred (experiences) and organisational (cost-effectiveness). An extensive literature search using MEDLINE (PubMed) EMBASE, and CINAHL was conducted. Primary studies with adults in primary care with both quantitative (comparison with physician-led care and cost-effectiveness) and qualitative (patient experiences of nurse-led care) methodologies from 2003 until June 2018. All studies were appraised using the Cochrane Collaboration's tool for assessing risk of bias. The appraisal involved evaluation of the degree of risk of bias in selection, performance, detection, attrition and reporting. Because of the complexity of multiple outcomes (quantitative and qualitative) a narrative synthesis was undertaken. The search generated 18 published studies that met our eligibility criteria. Three randomized controlled trials and one historical control trial found statistically significant differences in glycemic levels in favour of the nurse-led interventions. Two cluster randomized trials, two randomized pragmatic trials and two randomized controlled trials found no differences between groups. The three open-label studies found statistically significant improvements in HbA1c levels. The audit identified that more patients had lower HbA1c levels after the initiation of a nurse-led intervention. Three randomized controlled trials found significant improvements in biological outcomes and one did not. The four studies measuring cost-effectiveness found the nurse-led intervention was associated with less costs. Four studies examined patient satisfaction with nurse-led care and found this was very good. This review which incorporated a broad range of studies to capture the complexity of nurse-led interventions has identified that there is evolving evidence that nurse-led interventions for community treatment of diabetes may be more clinically effective than usual physician-led care. [ABSTRACT FROM AUTHOR]